which function I can use?

Hello,

There are two matrices.

I want to calculate the p value of each row.

>> x

x =

    0.3192    1.0933   -0.0068   -1.0891   -1.4916
    0.3129    1.1093    1.5326    0.0326   -0.7423
   -0.8649   -0.8637   -0.7697    0.5525   -1.0616
   -0.0301    0.0774    0.3714    1.1006    2.3505
   -0.1649   -1.2141   -0.2256    1.5442   -0.6156
    0.6277   -1.1135    1.1174    0.0859    0.7481

>> y

y =

   -0.1924   -0.1774   -0.8045   -1.1480   -0.0825    0.1001    1.7119
    0.8886   -0.1961    0.6966    0.1049   -1.9330   -0.5445   -0.1941
   -0.7648    1.4193    0.8351    0.7223   -0.4390    0.3035   -2.1384
   -1.4023    0.2916   -0.2437    2.5855   -1.7947   -0.6003   -0.8396
   -1.4224    0.1978    0.2157   -0.6669    0.8404    0.4900    1.3546
    0.4882    1.5877   -1.1658    0.1873   -0.8880    0.7394   -1.0722

    function [P] = p_row(x,y)
    x1=size(x,1);
    y1=size(y,1);
    for i=1:x1
    if lillietest(x(i,:)) == 1
        P(i)=ranksum(x(i,:),y(i,:));
    elseif lillietest(y(i,:)) == 1
        P(i)=ranksum(x(i,:),y(i,:));
    elseif vartest2(x(i,:),y(i,:)) == 1
        P(i)=ranksum(x(i,:),y(i,:));
    else
       [tmpH,tmpP]=ttest2(x(i,:),y(i,:));
       P(i)=tmpP(1);

    end

    end

>> p=p_row(x,y)

p =

    0.7961    0.2773    0.1490    0.1877    0.6335    0.6010

Answer 1

[yue li]

Hello, Star Strider,

Thank you for your response!

1. I want to calculate the p value of each row in two matrics.

2. I prefer to use ttest2().

3. If the values in each row are not normal distribution, I have to use ranksum().

4. If the values of each row in two matrics are not equal variance, I have to use ranksum().

My output is shown above.

How can I only have the last line of output?

Thank you so much for your help!

Yue